Boundary integral formulation for 2D and 3D thermal problems exhibiting a linearly varying stochastic conductivity (Q1914983)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Boundary integral formulation for 2D and 3D thermal problems exhibiting a linearly varying stochastic conductivity
scientific article

    Statements

    Boundary integral formulation for 2D and 3D thermal problems exhibiting a linearly varying stochastic conductivity (English)
    0 references
    0 references
    0 references
    0 references
    23 March 1997
    0 references
    This work examines steady-state heat conduction in a stochastic heterogeneous medium, where the thermal conductivity varies linearly along one direction, and its slope is a sum of a constant part and a zero-mean random part. As the first step, the governing Laplace equation is solved using a coordinate transformation of independent spatial variables, and the exact Green functions in both two and three dimensions are obtained for a linearly varying conductivity profile. In addition, a boundary integral equation statement in which the Green functions appear as kernels is concurrently obtained. Next, material stochasticity is introduced, and the perturbation approach is employed for deriving the mean value and covariance of the Green functions up to the second order. Perturbations are also used to solve the discretized boundary integral equation, so that mean vectors and covariance matrices can be obtained for the response, temperature and heat flux. An example involving steady-state temperature distribution in a block where conductivity varies due to a buried heat source illustrates the method. Finally, comparisons are made with Monte Carlo simulations.
    0 references
    0 references
    0 references
    0 references
    0 references
    stochastic heterogeneous medium
    0 references
    zero-mean random part
    0 references
    Laplace equation
    0 references
    coordinate transformation
    0 references
    Green functions
    0 references
    perturbation approach
    0 references
    mean value
    0 references
    covariance
    0 references